When did you first begin to investigate this particular area of lexicography? How would you define this pursuit?

My old weblog about all things urban--World New York, kind of a mix of what Gothamist and The Morning News are now--had a sidebar where I'd note interesting words I'd come across, especially those in other languages. This was well before I got a job writing dictionaries with Oxford University Press. Some people mentioned that they liked the sidebar words more than they did the regular content. I considered that, and eventually agreed that sidebar was indeed more interesting.

So I took the methods that I had been using to find stories about New York City, or any major city, and I applied them to hunting for words. I had been using Excite's keyword-based automated search as well as custom Moreover feeds and a few other methods. Starting with the results from a preset mish-mash of included and excluded words, names, and expressions, I'd go through hundreds of articles a day looking for just the right quotes on the subject of the Big Apple or city life--not necessarily the most representative, but the most striking, the best written, the ones with the most humor or insight.

Doing what is called a diachronic dictionary--also known as a historical dictionary, such as the Oxford English Dictionary--where each entry is supported by citations, is very similar to what I was doing on my old blog. You've got to dig or read a great deal to find a quote that really helps explain the history of a word, its nuances, the forms of expression it typically is found in, that subtle context that might benefit a learner or a non-native speaker.

A diachronic dictionary, then, was a natural for the new web site, now called Double-Tongued Word Wrester (though usually just called Double-Tongued aloud). The book is based on and is complementary to that site.

Which subgenre--for example, political, sexual, drug-related--is your favorite?

I have no favorites here, but in the broader categories of registers of language, jargon is the least easy to love. It's very boring. Where it's not boring, it's usually slang that's been mislabeled. Few people want to learn the names of 15 different types of pliers. Professionally, though, jargon is appealing because its boredom has meant that it has been relatively untouched and when it has been dealt with, it's usually been done badly.

Out of the various subjects, sexual slang is the hardest to treat evenly. On the one hand, I want to be as academic and adult about it as possible, so there's no room for giggling or guffawing. On the other hand, I inevitably end up supporting such entries with horrific citations of human depravity from alt.tasteless or similar newsgroups. See "squick" for an example. Adding to the discord, Google AdSense usually refuses to put ads on pages with sexual entries, which leaves Yahoo to show its dreadful acne cream advertisements.

Which ones in The Official Dictionary of Unofficial English are your favorite?

I stole this answer from my colleague, Erin McKean, head of US Dictionaries for Oxford University Press: I have no favorites. That'd be like preferring one of my children over another.

I will say that there are two types of entries I like a great deal: those that require a lot of detective work and those that show a word migrating into English from another language. The two types are often the same entries. "Chones" is one in the book. It's originally from Mexican Spanish and means "panties" or "underpants." It's so far into American English, and so widespread, that many English-only speakers use it without any awareness of its origins. This is best demonstrated by the varying pronunciations and spellings: "choners" and "chonies" are common.

On the web site, two similar entries are "fladry" and "wila." "Fladry" comes from the Polish and refers to flags used to contain or exclude wild animals. I had to call on European experts in wild wolf management and on professors of the Polish language to help me nail that one down.

"Wila" is even more interesting. Its transformations are like this: it started as "huilotl," Náhuatl for "dove," then became "güilota" or "huilota" also meaning "dove," which became "kite" (the kind you fly), shortened as "güila" or "huila" in Mexican Spanish. Then, relatively recently when Mexicans began to be imprisoned in US jails, they came across the American English prison slang "kite," meaning "a surreptitious note (sent by or between prisoners)." So they made "güila" or "huila" a calque--a loan translation--for the American English slang "kite," perfectly mirroring in Spanish the relationship of the standard and slang meanings already in English. Then the Anglophones got of ahold of it and began spelling it "wila."

Which words are the hardest to track down the origin of?

The common words, especially new meanings of old words or sense alterations of existing definitions. Think of trying to research "a put" as used in financial markets. It can eat a week if you let it. I don't have that kind of time, so I try to avoid such words. Corpora can be useful in this work, however. I've just returned from the United Kingdom where Patrick Hanks was lecturing on how best to use corpora analysis for creating dictionaries. It's really very interesting and should be leading to a lot of innovations, not the least making it easier to find those subtle changes that are otherwise hard to dig for. I'm not sure I can apply these lessons to slang or my web site, though. As Hanks kept reminding me, neologisms and slang--my main interests as a lexicographer--are boundary cases that require vast corpora even larger than the one-billion-word corpus that Oxford University Press has put together. The thing that would make it easiest is if the entirety of Google's various indexes (web, Groups, print, etc.) could be made into a makeshift corpus, one without tagging but with information on frequency, proximity, collocates, and with some kind of efficient lemmatizer and stemmer.

What's on your watchlist right now?

Always on the lookout for more new military slang. People seem to like it and it's usually pretty easy to verify. For the book, I used Flickr to find people who served in the military in Iraq by looking for users who had posted photos of the country. With their help, I was able to cull a few words from the book that the military folks insisted were invented by journalists. It's an inaccurate way of doing research--even if they were all in agreement about a word being "fake," they could be wrong--but I have a budget of $0 so I use what free methods I can.

What contribution, if any, has the internet--and perhaps blogging in particular--played in the creation of new words?

The main thing is that it has made in-groups both more visible and made their boundaries more well defined. For example, there's a crapload of World of Warcraft slang, which would never have existed nor would I have known about it without the Internet. But a cursory investigation of its more common forms shows that it almost never leaves WoW. At best, it is sometimes found in other, similar games. So it's not necessarily something I need to put on the record because even though its users might number in the millions, it's still a niche group. It has the same utility outside of WoW as airplane manufacturing jargon does outside of Boeing or Airbus: very little. Digital resources (mostly the Internet) made that whole process possible: the creation of the community (a.k.a the in-group), the creation of its own lexicon, the bringing of that lexicon to my attention, my investigation of its survivability outside the WoW word crèche, etc.

As for the creation of new words, I'd say the Internet has no special status as a neologism factory. As with any broad-reaching technology or medium (because it is both, after all), it has certainly generated plenty of terms that are specific to it, but there's simply nothing remarkable about the words created there: there's not an unexpected volume per capita, nor a particular level of genius, nor are words being created in new ways. There's also no evidence that Internet-originating neologisms last more often, last longer, or even that they leave the in-group and go into standard American English faster than neologisms from other modes of communication. The only characteristic of the Internet that might be significant is that their words have the potential to propagate faster and wider: but few do. Just as elsewhere, most Internet-originating neologisms die like mayflies or stay restricted to relatively small groups.

I frequently see sissy-fighting in blogs about who coined what word and when. Credit for coinage on the Internet is really a useless, fruitless honor, same as it is offline. The act of coinage is far too common to think that a single instance of it is noteworthy in any way. More useful and more interesting is the underlying language from which the word stems, its transformations, its uses new and old.